Re: The art of life and death: 14 year follow-up analyses of associations between arts engagement and mortality in the English Longitudinal Study of Ageing
CCBYOpen access
Rapid response to:
ResearchChristmas 2019: Express Yourself
The art of life and death: 14 year follow-up analyses of associations between arts engagement and mortality in the English Longitudinal Study of Ageing
Re: The art of life and death: 14 year follow-up analyses of associations between arts engagement and mortality in the English Longitudinal Study of Ageing
Dear Editor
We read with great interest the study by Fancourt et al. which concluded that arts engagement is associated with a lower mortality in older patients. There are, however, several limitations that could have confounded the interpretation of this study.
First, engagement in the arts was self-reported as were the covariates used for the sensitivity analyses. Self-reporting is associated with a significant risk of bias. It is very plausible that healthy individuals were more likely to remember and report prior engagement in the arts. Furthermore, individuals with chronic disease are less likely to engage in the arts or accurately remember their prior engagement in such activities. Despite the authors rigorously controlling for confounding variables, the nature of observational data dictates that it is impossible to control for all confounders and the E value may have provided false hope that confounders were not present.
Second, it is suggested that a plausible mechanism for the decrease in death is the increased social capital and subsequent successful aging that comes with engaging in the arts. However, the decrease in mortality was observed in those reporting arts engagement 1-2 times per year only at the study baseline and it is unclear if these mechanisms can reduce mortality with such infrequent engagement in the arts.
Third, this study provides no information on continued arts engagement and mortality as arts engagement was only assessed at one time point in the study. What would be more interesting and clinically significant would be a trial testing arts engagement in a randomized fashion as this would eliminate the significant bias observed in the current study.
Rapid Response:
Re: The art of life and death: 14 year follow-up analyses of associations between arts engagement and mortality in the English Longitudinal Study of Ageing
Dear Editor
We read with great interest the study by Fancourt et al. which concluded that arts engagement is associated with a lower mortality in older patients. There are, however, several limitations that could have confounded the interpretation of this study.
First, engagement in the arts was self-reported as were the covariates used for the sensitivity analyses. Self-reporting is associated with a significant risk of bias. It is very plausible that healthy individuals were more likely to remember and report prior engagement in the arts. Furthermore, individuals with chronic disease are less likely to engage in the arts or accurately remember their prior engagement in such activities. Despite the authors rigorously controlling for confounding variables, the nature of observational data dictates that it is impossible to control for all confounders and the E value may have provided false hope that confounders were not present.
Second, it is suggested that a plausible mechanism for the decrease in death is the increased social capital and subsequent successful aging that comes with engaging in the arts. However, the decrease in mortality was observed in those reporting arts engagement 1-2 times per year only at the study baseline and it is unclear if these mechanisms can reduce mortality with such infrequent engagement in the arts.
Third, this study provides no information on continued arts engagement and mortality as arts engagement was only assessed at one time point in the study. What would be more interesting and clinically significant would be a trial testing arts engagement in a randomized fashion as this would eliminate the significant bias observed in the current study.
Competing interests: No competing interests